[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: "load" on Windows
From: |
Eli Zaretskii |
Subject: |
Re: "load" on Windows |
Date: |
Sat, 05 Oct 2013 16:46:29 +0300 |
> Date: Sat, 05 Oct 2013 16:34:11 +0300
> From: Eli Zaretskii <address@hidden>
> Cc: address@hidden
>
> Paul, if this limitation is deliberate, I suggest to document it where
> we explain the arguments of gmk_add_function.
One other important thing that doesn't seem to be covered in the
manual is the requirements from the function that implements the
Make function added by a loaded object. If the reader looks at
gnumake.h, she will see that the function's signature should be like
this:
char *(*func)(const char *nm, int argc, char **argv)
But this had better be documented explicitly in the manual.
More importantly, we never say that the 'char *' value returned by the
extending function must either be a pointer to a buffer allocated by
calling gmk_alloc, or NULL; any other kind of pointer will cause
unexpected results or crashes, because Make will try to free any
non-NULL pointer returned by the extension. I think it is quite
important to describe this in the manual.
(I can make these changes in documentation, if you agree.)
- "load" on Windows, Gisle Vanem, 2013/10/03
- Re: "load" on Windows, Eli Zaretskii, 2013/10/03
- Re: "load" on Windows, Gisle Vanem, 2013/10/03
- Re: "load" on Windows, Eli Zaretskii, 2013/10/04
- Re: "load" on Windows, Eli Zaretskii, 2013/10/04
- Re: "load" on Windows, Gisle Vanem, 2013/10/05
- Re: "load" on Windows, Eli Zaretskii, 2013/10/05
- Re: "load" on Windows,
Eli Zaretskii <=
- Re: "load" on Windows, David Boyce, 2013/10/05
- Re: "load" on Windows, Eli Zaretskii, 2013/10/05
- Re: "load" on Windows, Paul Smith, 2013/10/05
- Re: "load" on Windows, Gisle Vanem, 2013/10/05